JP3402054B2 - Road white line recognition device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a road white line recognition apparatus for grasping a shape represented by a white line on a road on which an automobile is traveling, and more particularly to a road white line recognition apparatus using image processing that tends to take a long time. The present invention relates to a road white line recognition device capable of correcting a recognition delay according to image processing time while recognizing a white line.

[0002]

2. Description of the Related Art In recent years, a television camera has been installed in an automobile to grasp the shape of a road while traveling and to grasp the posture of the vehicle with respect to the road and use it for automatic traveling control of the automobile.
Techniques have been developed that are used for various warnings to drivers. For example, in the image information from the camera installed in the vehicle, the part with high lightness is set as the white line candidate point, and the point closest to the white line position (white line recognition information) obtained from the previous screen information is recognized as the white line. Then, a technique relating to road white line recognition, in which steering is performed based on the road white line, is disclosed in Japanese Patent Laid-Open No. 3-137798.

Further, in Japanese Patent Laid-Open No. 7-85249,
Based on the image information, search the brightness change in the horizontal direction in the white line search area, judge that the part where the brightness change is extremely small is the road part, and change the brightness next to the road part where the brightness change is very small. A technique for recognizing a large line as a white line is disclosed. Such a road white line recognition device includes a camera (imaging unit) 11 and an ECU (electronic control unit) 15A for processing image information from the camera 11, and outputs white line data 19A as shown in FIG. In this case, as shown in FIG. 11, for example, first, the camera 11 acquires an image (actual image) (step C10), and the ECU 15A processes this image (step C20). Then, as shown in FIG. 12, the image (original image) 121 is converted into a two-dimensional image 122 in plan view (step C30),
Road white lines 121A and 121B of the original image 121 are converted into road white lines 122A and 122B of the planar view image 122, and then white line recognition is performed.

By the way, when the vehicle rides on the bumps and the like of the road while the vehicle is running, the posture of the vehicle body changes, and the direction of the vehicle-mounted camera also changes vertically or horizontally accordingly. Therefore, the image captured by the vehicle-mounted camera also changes vertically or horizontally, and the vehicle running must recognize the road shape (white line of the road) in consideration of the influence of such road irregularities.

Therefore, as disclosed in Japanese Unexamined Patent Publication No. 6-124398, it is determined whether or not the wheels ride on uneven roads or the like based on the pitch of the vehicle body by detecting the pitch and the amount of movement of the vehicle body while traveling. If the wheel rides on the unevenness of the road, it does not recognize the shape of the road based on the in-vehicle camera captured at this time. A technique for recognizing a road shape by adding the amount of movement of a vehicle is disclosed.

[0006]

The image processing for recognizing the road white line (road shape) as described above has a low processing speed and attempts to perform some control based on the recognition of the road white line. However, it may not be the proper control. Therefore, higher-speed image processing is desired, but an expensive and high-speed electronic control unit (ECU) must be used to perform high-speed image processing, resulting in a large cost burden. .

The present invention was devised in view of the above-mentioned problems, and provides a road white line recognition method capable of accurately recognizing a road white line so that appropriate control can be performed at low cost. With the goal.

[0008]

Therefore, the road white line recognition apparatus of the present invention according to claim 1 is mounted on a vehicle, and image pickup means for photographing a road in the traveling direction of the vehicle and outputting the photographed image, original for the image information captured by the image pickup means
An image information processing means for recognizing a white line on the road by performing image processing including a processing for converting an image into a planar view image, a vehicle speed detecting means for detecting a vehicle speed of the vehicle, and a yaw rate of the vehicle are provided. A yaw rate detecting means for detecting, a calculating means for calculating a movement amount and an attitude change amount of the vehicle based on the vehicle speed detected by the vehicle speed detecting means and the yaw rate detected by the yaw rate detecting means, and the imaging means. The amount of movement and posture change of the vehicle within the image processing time, which is calculated by the calculating unit during the image processing time from the time when the image is picked up to the time when the road white line recognition is completed by the image processing. It is characterized in that it is provided with a correcting means for correcting the road white line recognition information recognized by the image information processing means.

According to a second aspect of the present invention, there is provided a road white line recognition apparatus according to the first aspect, wherein the image information processing means is provided.
Is the image information captured by the imaging means in the first period by performing image processing to recognize the white line該道path, the calculated
A means for determining the first speed based on the vehicle speed detected by the vehicle speed detection means and the yaw rate detected by the yaw rate detection means.
The movement amount and the posture change amount of the vehicle are calculated in a second cycle shorter than the cycle , and the correction means completes the recognition of the road white line.
The amount of movement of the vehicle according to the above image processing time at the time and
After correcting the road white line recognition based on the posture change amount,
During the correction, until the completion of recognition of the road white line in the cycle
It is characterized in that the road white line recognition information is further corrected based on the movement amount and the posture change amount of the vehicle from the point to the present time.

According to a third aspect of the present invention, in the road white line recognizing device of the first aspect, it is determined whether or not the posture change amount of the vehicle is larger than a preset determination value. It is characterized in that it comprises a determining means, and the correcting means performs the above correction when the determining means determines that the posture change amount is larger than the determination value.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of the present invention will be described below with reference to the drawings. FIGS. 1 to 9 show a road white line recognition apparatus as an embodiment of the present invention. It should be noted that this device can be provided in an automobile. First, the purpose of using the road white line recognition by this device will be explained.To recognize the road white line, the driver can be guided or warned so that the vehicle runs along the driving lane so that the vehicle does not deviate from the driving lane. Furthermore, it is necessary for automatic steering control that allows the vehicle to be automatically steered along the traveling lane.

For example, FIG. 2 is a schematic diagram showing an automatic steering device. In the vehicle 1, a steering actuator 7 for steering the steered wheels 2 and an image pickup means for picking up an image of a road condition in front of the vehicle. Camera 11 and this camera 1
Image information processing means 15 for processing image information from 1;
The steering actuator 7 is provided with a controller 5 as a control means for setting a control signal.

Based on the image captured by the camera 11, the steered wheels 2 can be automatically steered without the driver having to operate the steering wheel 6. The target steering angle is set as a control amount for controlling the steered wheels 2 based on the information from the processing means 15, and the steering actuator 7 is operated according to the set amount of the target steering angle. is there.

Although this apparatus has such utility, control based on road white line recognition (white line recognition corresponding control) such as guidance to a driver or automatic steering is originally controlled from the time when image information is obtained. There should be no time difference. However, in reality, since the image processing speed of the vehicle-mounted image processing apparatus is limited, the road white line recognition information that is image-processed and output will be the information before the image processing time T. Therefore, the present device is configured to calculate the movement of the vehicle during the image processing time T and correct the image-processed road white line recognition information based on the calculation result.

FIG. 1 is a block diagram schematically showing the construction of such an apparatus, but as shown in the figure, the apparatus has a camera 11 as an image pickup means and a vehicle speed for detecting a vehicle speed V of the vehicle. A vehicle speed sensor 12 as a detecting means, a yaw rate sensor 13 as a yaw rate detecting means for detecting a yaw rate ψ'generated in the vehicle, and an ECU (electronic control unit) 14 are provided.

The camera 11 is mounted on an automobile (hereinafter referred to as a vehicle), photographs a road in the traveling direction of the vehicle, and outputs the photographed image to the ECU 14. The ECU 14 is based on a function (image information processing means) 15 for performing image processing on image information picked up by the camera 11, a vehicle speed V detected by the vehicle speed sensor 12, and a yaw rate ψ ′ detected by the yaw rate sensor 13. A function (calculating means) 16 for calculating the movement amount and the attitude change amount of the vehicle (simply referred to as the attitude change) and the calculated attitude change amount of the vehicle are compared with a preset determination value for determination. Function (determination means)
17 and a function (correction means) 18 for correcting the road white line position (road white line recognition information) recognized by the image information processing means 15.

Then, the image information relating to the road white line recognition information is output at a cycle T1 corresponding to a processing cycle T1 by the image information processing means 15 which will be described later. Thus, the control is performed in the cycle T1 according to the road white line recognition information thus output. The image information processing means 15 performs image processing on the image information captured by the camera 11. In this processing,
In the camera 11, the perspective image (original image) 2 as shown in FIG.
While 1 is captured, it is converted into a two-dimensional image 22 in plan view as shown in FIG. Therefore, the road white lines on the image are recognized after the original images 21A and 21B are converted into the planar view images 22A and 22B. Since the image processing by the image information processing unit 15 takes a certain processing time T, the image information processing unit 15 periodically performs the image processing at a cycle T1 (≈T) corresponding to the processing time T.

Based on the detected vehicle speed V and yaw rate ψ ', the calculating means 16 calculates the vehicle front-rear movement amount X, the vehicle left-right movement amount Y, and the vehicle posture change amount. The yaw angle change amount Ψ is calculated. The movement amounts X and Y of the vehicle and the posture change amount Ψ calculated by the calculating means 16 are as described above.
The purpose of calculation is to correct the road white line recognition information by the correction means 18.

That is, as described above, since the processing time T by the image information processing means 15 is long, if the vehicle moves or changes its posture during this time, the road white line recognition output from the image information processing means 15 will be recognized. The information is different from that at the time of image information acquisition. Therefore, the movement amounts X and Y of the vehicle and the attitude change amount Ψ from the time when the image information is acquired to the time when the road white line recognition information is output are calculated, and the road white line recognition information output by the correction unit 18 is calculated as follows. It is intended to output more accurate road white line recognition information by correcting the vehicle movement amounts X and Y and the posture change amount Ψ.

The calculation of the movement amounts X and Y and the posture change amount Ψ will be described. The front-back movement amount X, the left-right movement amount Y, and the yaw angle change amount Ψ are the front-back movement amount Δ per minute time Δt.
x, the amount of lateral movement Δy, and the yaw angle ψ are obtained by time integration. The amount of longitudinal movement Δx is the product of vehicle speed V and time Δt, and the amount of lateral movement Δy is the amount of longitudinal movement Δx and the yaw angle change per minute time Δt. The yaw angle ψ can be calculated as the product of the yaw rate Δψ ′ per minute time Δt and the time Δt, and the yaw angle ψ can be calculated as the product of the following expressions (1) to (3), respectively.
Can be calculated by X = ∫Δxdt = ∫ (V · Δt) dt ··· (1) Y = ∫Δydt = ∫ (Δx · Δψ) dt ··· (2) Ψ = ∫ψdt = ∫ (Δψ ′ · Δt) dt (3) The forward / backward movement amount X and the left / right movement amount Y by this calculation means 16
The yaw angle change amount Ψ is calculated at the time when the image information is captured by the camera 11 (at this time, the image information processing means 15 is used.
Since there is no problem even if this image information is input at the time, the image information processing means 15 completes the image processing and outputs this processing information. The calculation means 1
In 6, the calculation is started at the time of inputting the image information in the cycle T1, and thereafter, the calculation is performed by periodically taking in the detection information of the vehicle speed V and the yaw rate ψ'at a predetermined cycle T2 shorter than the image processing cycle. The calculation is completed when the image processing information is output.

FIG. 6 is a time chart showing an example of calculating the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ. Here, the image processing cycle T1 is the vehicle speed V and the yaw rate ψ '.
Of the detection information acquisition period (calculation period) T2 times m (T
1 = m · T2), that is, in the case where the detection information of the vehicle speed V and the yaw rate ψ ′ is set to be acquired m times in one image processing cycle, a certain image processing cycle [nth and (n + 1) ) Second image processing cycle], an example of calculation of the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ will be described.

In the n-th image processing cycle, first, a camera image is acquired at a time point (t _n-1 ) and then at this time point (t _n-1 ).
At the time point (t _n ) after the lapse of time T1 from ( _n−1 ), a planar view image is obtained by image processing. During this period, the front-rear movement amount Δx _ni , the left-right movement amount Δy _ni , and the yaw angle ψ _ni per time T2 are calculated while taking in the vehicle speed V and the yaw rate ψ ′ in the cycle T2, and these are respectively calculated by the following equation (4)
The integration is performed by (6) to calculate the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ from the time point (t _n-1 ) to the time point (t _n ). The expressions (4) to (6) correspond to the expressions (1) to (3). X _n = ΣΔx _nk, where k = _{1 to m} (4) Y _n = ΣΔy _nk, where k = _{1 to m} (5) Ψ _n = Σψ _nk, However, k = _{1 to m} (6) In this way, the calculated forward / backward movement amount X _n , left / right movement amount Y _n , and yaw angle change amount Ψ _n are processed in the n-th image processing cycle. It is used to correct the image information.

In FIG. 6, the process of calculating the front-back movement amount X, the left-right movement amount Y, and the yaw angle change amount Ψ in the n-th image processing cycle is shown by equations (4A) to (6A). There is. X _n = ΣΔx _nk, where k = 1 to i (4A) Y _n = ΣΔy _nk, where k = 1 to i (5A) Ψ _n = Σψ _nk, However, k = 1 to i (6A) Further, the calculation result for the (n-1) th image processing cycle is as shown in Expressions (4B) to (6B). The back-and-forth movement amount X _n-1 , the left _- right movement amount Y _n-1 , and the yaw angle change amount Ψ _n-1 are used for correcting the image information processed in the (n-1) th image processing cycle. X _n-1 = ΣΔx _{(n-1) k,} where k = _{1 to m} ... (4B) Y _n-1 = ΣΔy _{(n-1) k,} where k = _{1 to m}・ ・ ・ ・ ・ (5B) Ψ _n-1 = Σψ _{(n-1) k,} where k = 1 to m (6B) and calculation for the (n + 1) th image processing cycle The process is as in equations (4C) to (6C), and the calculated forward and backward movement amount X _{n + 1} , left and right movement amount Y _{n + 1} , and yaw angle change amount Ψ _{n + 1} are the n + 1-th image processing. It is used to correct image information processed in cycles. X _{n + 1} = ΣΔx _{(n + 1) k,} where k = _{1 to m} ... (4C) Y _{n + 1} = ΣΔy _{(n + 1) k,} where k = _{1 to m} (5C) Ψ _{n + 1} = Σψ _{(n + 1) k,} where k = _{1 to m} (6C) The determination means 17 is image processing information relating to road white line recognition. Is to determine whether or not the correction is required. It is based on the idea that if there is little movement or change in posture of the vehicle during the processing of the image information by the image information processing means 15, no correction is necessary. In particular, for vehicle movement during image processing, move the vehicle along the lane (ie, along the road white line).
This is not a serious problem in running the vehicle, but becomes a serious problem when the vehicle undergoes a certain attitude change.
Therefore, the determination means 17 compares the posture change amount Ψ calculated by the calculation means 16 with the determination value Ψ _0, and calculates by the calculation means 16 only when the posture change amount Ψ becomes equal to or larger than the determination value Ψ _0. Based on the result, it is determined that the image processing information needs to be corrected.

In the correcting means 18, when the judging means 17 judges that the image processing information needs to be corrected, the calculating means 16
Based on the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ calculated in step 1, the obtained image information is corrected so as to correspond to the current position and orientation of the vehicle. For example, in the image information processing means 15, when viewed in plan as shown in FIG.
Although converted into the three-dimensional image 22, during the processing of this image information, the vehicle changes from the position and orientation indicated by reference numeral 30 in FIG. 4 to the position and orientation indicated by reference numeral 31 in FIG. In this case, the correction means 18 corrects the image information centered on the current position and posture of the vehicle (see reference numeral 31) as shown in FIG.

Therefore, in the image 23 corrected by the correction means 18 (corrected two-dimensional image in plan view), the position of the road white line is changed from the positions indicated by 22A and 22B in FIG.
Are corrected to the positions indicated by 23A and 23B. Of course, if the determination means 17 does not determine that the image processing information needs to be corrected, such correction is not performed. In this way, the image information relating to the road white line recognition information is appropriately corrected by the correction means 18 and is output.
As described above, it is equal to the image processing cycle T1 by the image information processing means 15.

Since the road white line recognition apparatus as an embodiment of the present invention is configured as described above, the road white line recognition processing is performed as shown in the flowcharts of FIGS. 7 and 8, for example. That is, FIG. 7 is the main routine, and steps A10 to A50 are repeated in the cycle T1. That is, first, an image as shown in FIG. 3 is acquired by the camera (step A10), and then the image information processing unit 15 performs a conversion process into a two-dimensional image in plan view as shown in FIG. 4 (step S10). A20) and white line data are obtained (step A30). Then, the white line data is corrected (step A40), and the road white line recognition process of the current cycle ends.

FIG. 8 is a routine for correcting the white line data. This routine is started in response to the start of the image information processing, and the cycle T2 is shorter than the image information processing. ~ Each process of step B70 is repeatedly performed. That is, first, the sensor data of the vehicle speed V and the yaw rate ψ'is acquired (step B1).
0), based on this, as shown in FIG. 6, the movement amount, that is, the front-back movement amount X and the left-right movement amount Y are integrated (step B
20), the posture change (posture change amount), that is, the yaw angle change amount Ψ is integrated (step B30).

Since such a process (steps B10 to B30) returns after the judgment process of step B40, steps B10 to B30 cycle T until the image processing by the image information processing means 15 is completed.
Repeated at 2. Then, when the image processing is completed, the attitude change amount obtained during this time, i.e., by comparing the yaw angle variation [psi and the determination value [psi ₀ (step B50), if the yaw angle variation [psi determination value [psi ₀ or more, As the image correction information, the integrated result of the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ is output (step B60). Then, the integrated values of the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ are cleared (step B70), and the calculation of the correction information of this cycle is completed.

On the other hand, if the yaw angle change amount Ψ is not greater than or equal to the judgment value Ψ ₀ , the integrated values of the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ are not output without outputting the image correction information. After clearing (step B70), the calculation of the correction information of this cycle is completed. In this way, in the present apparatus, when the vehicle undergoes a certain posture change during the time T required for image processing, the movement of the vehicle within the image processing time T is calculated, and the calculation result is calculated. Since the road white line recognition information that has been image-processed by is corrected, it is possible to properly grasp the actual vehicle position and posture with respect to the road white line based on the image information, and various controls based on the road white line recognition (white line recognition support control ) Can be performed accurately.

Next, a modification of this embodiment will be described. In this modification, the calculation cycle T2 by the calculation means 16 is shorter than the processing cycle T1 by the image information processing means 15.
Then, the correction means 18 performs the correction, and the road white line recognition information (image information) is output. The white line recognition corresponding control is also performed in accordance with the road white line recognition information thus output. The control is performed at the cycle T2.

Therefore, the correction amount by the correction means 18 in this case is calculated as shown in FIG. FIG. 9 corresponds to FIG. 6 according to the present embodiment and is a time chart showing an example of calculation of the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ. When image processing is completed (t _n-1 ),
The time point (t _n ) and the time point (t _{n + 1} ) are the same as those in the embodiment of FIG. 6, but the time point (t _n-1 ) after the image processing is completed.
And time point (t _n ) or time point (t _n ) and time point (t _{n + 1} ).
Between and, the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ are calculated by the following equations (7) to (9).

The expressions (7) to (9) are (4) to (6).
Corresponds to. X _n = ΣΔx _{(n-1) k1} + ΣΔx _nk2 where k1 = _{1 to m} , k2 = 1 to i (7) Y _n = ΣΔy _{(n-1) k1} + ΣΔy _nk2 where k1 = _{1 to m} , k2 = 1 to i (8) Ψ _n = Σψ _{(n-1) k1} + Σψ _nk2 where k1 = _{1 to m} , k2 = 1 to i 9) Thus, the amount of movement of the vehicle body (that is, the amount of forward / backward movement X _n , the amount of lateral movement Y _n ) and the amount of change in the posture of the vehicle body (that is, the amount of yaw angle change Ψ) that is shorter than the processing cycle T1 by the image information processing unit 15 _The corrected road white line recognition information (image information) is output at the calculation period T2 of _n ), and by performing the white line recognition corresponding control, even if the image processing period is long, there is an advantage that finer control can be performed. is there.

In this case, the correction procedure in the road white line recognition is also different from that of the embodiment shown in FIGS. That is, the step of “white line data correction” of step A40 in the flowchart (FIG. 7) of the road white line recognition routine performed in the cycle T1 is deleted,
Steps A10 to A30 are repeated in the cycle T1. Therefore, in this road white line recognition processing, the white line data that is not corrected in the cycle T1 is output (step A30).
Only do.

Then, in the flow chart (FIG. 8) of the correction of the road white line recognition performed in the cycle T2, the step of "image processing end judgment" of step B40 is deleted,
"Image processing end determination" between steps B60 and B70
Step (this is referred to as step B65) is added. Further, in step B60 of FIG. 8, the image correction information is output, but in step B60 of this modified example, the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change obtained in steps B20 and B30 of this cycle. Road white line recognition routine (routine corresponding to FIG. 7) based on the integration result of the quantity Ψ
The image information obtained in step A30 is corrected and output as corrected image information. That is, the corrected image information is output for each cycle T2 shorter than the cycle T1.

Therefore, in such a routine configuration, step B30 of FIG. 8 is followed by step B50, and therefore step B50 is step B30 of this cycle.
The amount of change in posture obtained in step 1, that is, the amount of change in yaw angle Ψ is compared with the judgment value Ψ _0, and this judgment is also made in the cycle T2. If the yaw angle change amount Ψ is greater than or equal to the determination value Ψ ₀ , the process proceeds to step B60, and the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ obtained in steps B20 and B30 of this cycle. The image information is corrected based on the integration result, and this is output as corrected image information. Therefore, the corrected image information is also output at every cycle T2 shorter than the cycle T1.

If "NO" in the step B50, directly, and if "YES" in the step B50, step B6.
After 0, the process proceeds to the image processing end determination step of step B65, where if it is determined that the image processing is completed,
In step B70, the integrated values of the forward / backward movement amount X, the left / right movement amount Y, and the yaw angle change amount Ψ are cleared. If the image processing is not completed, the process returns from step B65.

As a result, the integrated value is cleared every time image processing is completed (hence, every cycle T1). In the above-described embodiment, since the correction is performed only when the vehicle makes a certain posture change or more during the time T required for the image processing, the correction frequency can be suppressed to a low level. The correction may be always performed except for such a condition.

[0038]

As described above in detail, according to the road white line recognition apparatus of the present invention as set forth in claim 1, the captured image information is obtained.
Against when the vehicle makes a posture change between that image processing including processing for converting the original image into a plan view image lines Tsu name it is, calculates the movement of the vehicle in the image processing time, Since the image processed road white line recognition information is corrected based on this calculation result, the actual vehicle position and attitude with respect to the road white line can be appropriately grasped based on the image information, and the road white line can be accurately recognized with respect to the own vehicle. Therefore, various controls based on road white line recognition can be accurately performed.

According to the road white line recognizing device of the present invention as defined in claim 2 , in addition to the above effects, finer control is possible even if the image processing cycle is long. Claim 3
According to the described road white line recognition apparatus of the present invention , the correction frequency can be further suppressed, so that unnecessary correction processing can be avoided and the speed of information output can be increased.

[Brief description of drawings]

FIG. 1 is a block diagram schematically showing a configuration of a road white line recognition device as an embodiment of the present invention.

FIG. 2 is a schematic diagram of a vehicle showing an example of use of the road white line recognition device.

FIG. 3 is a diagram showing an example of image information imaged by an imaging unit of the present embodiment.

FIG. 4 is a diagram showing an example of image information processed by the image information processing means of the present embodiment.

FIG. 5 is a diagram showing an example of image information corrected by the correction means of the present embodiment.

FIG. 6 is a time chart for explaining correction contents by the correction means of the present embodiment.

FIG. 7 is a flowchart illustrating a procedure of road white line recognition according to the present embodiment.

FIG. 8 is a flowchart illustrating a correction procedure in road white line recognition according to the present embodiment.

FIG. 9 is a time chart for explaining correction contents by the correction means of the modified example of the present embodiment.

FIG. 10 is a block diagram schematically showing a configuration of a conventional road white line recognition device.

FIG. 11 is a flowchart illustrating a procedure of conventional road white line recognition.

FIG. 12 is a diagram illustrating image processing by a conventional road white line recognition device.

[Explanation of symbols]

1 vehicle 2 steering wheels 5 Controller as control means 6 steering wheel 7 Steering actuator 11 Camera as image capturing means 12 Vehicle speed sensor as vehicle speed detection means 13 Yaw rate sensor as means for detecting yaw rate 14 ECU (electronic control unit) 15 Image information processing means 16 Calculation means 17 Judgment means 18 Correction means

Front Page Continuation (72) Inventor Tadashi Sugawara 5-3-8, Shiba, Minato-ku, Tokyo Mitsubishi Motors Corporation (72) Inventor Norio Inoue 5-33-8, Shiba, Minato-ku, Tokyo Mitsubishi Motors Kogyo Co., Ltd. (56) Reference JP-A-3-265007 (JP, A) JP-A-3-276211 (JP, A) JP-A-9-161060 (JP, A) JP-B-6-10838 (JP , B2) (58) Fields surveyed (Int.Cl. ⁷ , DB name) G01B 11/00-11/30 G06T 7/00 G01C 21/00-21/36 H04N 7/18 G05D 1/02 G08G 1 / 16

Claims (3)

(57) [Claims] 1. A plane and the image pickup means, the original image for the image information captured by the imaging means mounted on a vehicle and outputs the image photographed by photographing the traveling direction of the road of the vehicle
An image information processing means for recognizing a white line on the road by performing image processing including a processing for converting into a visible image, a vehicle speed detecting means for detecting a vehicle speed of the vehicle, and a yaw rate for detecting a yaw rate of the vehicle. Detection means, calculation means for calculating the movement amount and posture change amount of the vehicle based on the vehicle speed detected by the vehicle speed detection means and the yaw rate detected by the yaw rate detection means; Based on the movement amount and the attitude change amount of the vehicle within the image processing time, which is calculated by the calculating unit during the image processing time from the time when the road white line recognition is completed by the image processing to the time when the road white line recognition is completed. A road white line recognition device comprising: a correction unit that corrects the road white line recognition information recognized by the image information processing unit. 2. The image information processing means recognizes a white line on the road by performing image processing on the image information picked up by the image pickup means in a first cycle, and the calculating means is the vehicle speed detecting means. Based on the vehicle speed detected by the yaw rate and the yaw rate detected by the yaw rate detection means, the movement amount and the attitude change amount of the vehicle are calculated in a second cycle shorter than the first cycle, and the correction means From after the correction of the road white line recognition based on the amount of movement and the attitude change amount of the vehicle according to the image processing time at the time of completion of the road white line recognition until the time when the road white line recognition of the next cycle is completed The road white line recognition device according to claim 1, wherein the road white line recognition information is further corrected based on the movement amount and the attitude change amount of the vehicle from the correction time point to the present time point. 3. A determination means for determining whether or not the posture change amount of the vehicle is larger than a preset determination value, the correction means includes the determination means for determining the posture change amount by the determination value. The road white line recognition device according to claim 1 or 2, wherein the correction is performed when it is determined that the road white line is larger than the above.